Tobias Scheer Université de Nice CNRS 6039 [email protected]

CLITE 1 Szeged, 19-21 April 1998

Vowel - Zero Alternations in Czech Prefixes (1) [e]-zero alternations in Czech prefixes +e -e beze-dný bezø-kv!tný vze-dmout vzø-hled pÍede-vším pÍedø-skok rozø-dmýchat roze-dmout roze-pÍít rozø-pÍahat

"without bottom/ without flowers" "blow up/ expression (face)" "before all/ test-jump ('before-jump')" "blow up/ fan" "strut/ remove"

(2) conditions on this alternation a. only C-final prefixes alternate do-hovoÍit *doe-hovoÍit "finir de parler" do-cela *doe-cela "tout à fait" do-hra *doe-hra "épilogue", NOMsg do-her *doe-her "épilogue", GENpl b. only √CC-initial stems provoke the alternation roz-dat *roze-dat "distribuer" pÍed-loni *pÍede-loni "il y a deux ans" bez-pe…í *beze-pe…í "sécurité" nad-hodnota *nade-hodnota "plus-value" (3) zero-forms are underlying a. corresponding prepositions have zero-forms in isolation bez, pÍed, nad, pod,... b. vocalized forms are the consequence of consonant-clusters like elsewhere in the language -CøC-V -CeC-ø -CeC-CV hudb-a NOM sg hudeb GEN pl hudební Adj kavárn-a NOM sg kaváren GEN pl kaváren-ský Adj GEN sg loket NOM sg loketní Adj lokt-u (4) identical √CC- behave contrastively cluster provoking provoking prefixal Vvs. prefixal zero √BR ode-BRat vs. bezø-BRadý √DR roze-DRat vs. rozø-DRobit vs. rozø-ŠLapat √ŠL vze-ŠLý √PÌ roze-PÌít vs. rozø-PÌahat

"take away/ beardless" "tear up/ crumble" "open (flower)/ crush" "spread around (tech)/ stretch out"

(5) corpus: exhaustive data from Ulbrich (1978) concerning 7 C-final prefixes (out of 11, ob- and the three "non-syllabic" items s-, z-, v- miss). Numeric overview: nb items nb items prefix +e -e gloss bez vz pÍed roz nad pod od sum TOTAL

16 11 16 80 5 26 41 195

39 20 48 295 33 74 253 762 957

"without" inchoative, "up" "before, in front of" inch., "disperse/ break into pieces" "over" "under" distantiational movement

(6) the secret must be found in the different status of stem-initial CC-clusters. (7) stem-initial CCs observed with a. prefixal -V only +e b. prefixal -ø only -e c. both mix +e only: 17 CCs

-e only: 38 CCs

ct, dn, dÍ, jm, lstn, mk, pn, ps, rv, Ív, sch, sr, šv, tn, vñ, zÍ, ñr

bl, bÍ, cl, cv, …l, fÁ, fr, hl, hm, hv, chl, chrchl, km, kr, kÍ, kv, mÁ, mr, pl, pt, sh, sv, šk, šn, šp, šr, tl, tr, tv, vd, vr, zbr, zp, zt, ñh, ñm, ñÁ, ñv

mix: 35 CCs br, …t, dm, dr, dv, hn, hr, hÍ, chv, jd, kd, kl, ml, mn, pj, pr, pÍ, sk, sl, sm, sn, sp, st, šl, št, tÍ, v…, vl, vÍ, vš, vz, zd, zl, zn, zv

TOTAL nb CC: 90 (8) A given root belongs to one and only one of these three groups.

(9) CC mix represented by how many items? underscored: CCs with significant numeric disproportion

items items items CC CC CC mix mix +e mix +e +e -e -e -e br 6 10 ml 5 14 št 1 14 …t 1 2 mn 2 6 tÍ 1 17 dm 3 2 pj 3 1 v… 1 2 3 14 pr 2 57 vl 1 24 dr dv 1 8 pÍ 8 20 vÍ 5 2 hn 3 1 sk 1 13 vš 2 1 hr 7 16 sl 8 6 vz 5 1 hÍ 3 2 sm 3 3 zd 4 3 chv 2 2 sn 1 1 zl 2 2 jd 2 2 sp 2 2 zn 6 2 kd 1 1 st 21 97 zv 4 4 šl 2 3 kl 1 53 (10) a. all CC mix with numeric disproportion are classical branching Onsets or s+Cs. b. almost all branching Onsets present a numeric disproportion. (11) Combien de racines par CC? nb racines CC br …t dm dr dv hn hr hÍ chv j(d) kd kl ml mn pj pr pÍ sk sl sm sn sp st šl št tÍ v… vl vÍ vš vz

+e 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 3 3 1 2 6 1 1 1 1 1 1 1 1

-e 7 1 1 11 3 1 5 1 2 1 1 11 3 1 1 10 7 5 2 3 1 1 24 2 1 4 1 6 1 1 1

CC zd zl zn zv bl bÍ cl cv …l fÁ fr hl hm hv chl chrchl km kr kÍ kv mÁ mr pl pt sh sv šk šÁ šp šr tl

nb racines +e 1 1 2 2 -

-e 2 1 2 3 2 4 1 1 2 1 1 6 1 1 1 1 2 10 5 2 7 4 22 1 1 5 7 1 1 2 2

nb racines CC tr tv vd vr zbr zp zt ñh ñm ñÁ ñv ct dn dÍ jm lstn mk pn ps rv Ív sch sr šv tn vñ zÍ ñr

+e 1 1 1 2 1 1 1 1 1 1 1 1 1 1 1 1 1

-e 7 1 1 5 1 1 1 1 1 1 1 -

(12) a. all +e CCs are represented by a single Root. b. many +e CCs are "exotic" in regard of ordinary IE branching Onsets: jm, dn, mk, kd, tn. (13) conclusion a. -e items are unmarked, both in overall count and within mixed CCs b. if +e items were to be disconsidered, root-initial CCs would coincide with what is classically known as a possible branching IE Onset. c. +e CCs are represented by a single Root. They are untypical IE word-initial clusters.

==> Something is wrong with +e CCs. (14) Solution: Root structure. +e CCs enclose a hidden zero. (15) +e √CC- Roots have alternating √CVC- forms -e √CC- Roots never do.

√CC√BR√DR√HR√HN√PR√SN√ŠL√ZD√DN-

+e Root two words from the same root a. b. ode-brat pf od-bírat roze-drat inf roz-deru pÍede-hra noun NOMsg her ode-hnat pf od-hán!t ode-prat inf od-peru beze-sný adj sen vze-šlý adj šel pode-zdít inf ze‹ beze-dný adj den

ipf 1Esg noun GENpl ipf 1Esg noun NOMsg pap masc sg noun NOMsg noun GENpl

-e Root non-related root c. bez-bradý roz-drobit od-hrabat roz-hn!vat vz-pruha pod-sn!ñník roz-šlapat od-zdola -

(16) +e Roots are always open. -e Roots are always closed by a third consonant. C2 is stem-final [√C1C2-]

(17)

C2 is part of the stem-initial cluster

=/C1__C2/

=/C1C2__/

√BR-

ode-B__R-at

vs.

bez-BRaD-ý

√DR-

roze-D__R-at

vs.

roz-DRoB-it

√HR-

pÍede-H__R-a

vs.

od-HRaB-at

√HN-

ode-H__N-at

vs.

roz-HN!V-at

√PR-

ode-P__R-at

vs.

vz-PRuH-a

√SN-

beze-S__N-ý

vs.

pod-SN!ð-ník

√ŠL-

vze-Š__L-ý

vs.

roz-ŠLaP-at

√ZD-

pode-Z__D-ít

vs.

od-ZDoL-a

√DN-

beze-D__N-ý

√/CCvC/ ==> +e √/CøC/ ==> -e

-

(18) given (17), all previous observations fall out naturally: a. "untypical" #CCs such as #jm etc. were observed with +e Roots. In fact, only C1 is root-initial, C2 being root-final: [C1C2] = /C1øC2]. b. +e CCs represent a single root and are marked because they come from jer-roots < √CjerC/ that are less frequent than non-jer roots. The number of roots with identical C1C2 is higher for /C1C2vC/ than for /C1vC2/. (19) (17) holds for ALL items and roots. Exhaustive data for the CCs involved in (15) (for all CCs, cf. Scheer 1996): Concordance CC - number & nature of the roots concerned CC

nb √

nb √"e

nb it

Roots representing the CC (one illustrations per root followed by the number of items representing it in brackets)

br

8

1+

9

roze-br-án (9)

7-

10

bez-bran-ný (1), bez-brad-ý (3), roz-brázd-it (1), roz-bre…-et (1), od-bru…-et (1), od-brebt-at (2), roz-broj (1)

1+

3

roze-dra-t (3)

11-

14

roz-drtit (1), roz-drbat (1), roz-drobit (3), roz-drolit (3), roz-druñovat (2), rozdráñdit (1), bez-dÍevý (1), roz-drápat (1), roz-drásat (1)

1+

7

ode-hrá-t (7)

5-

16

pod-hr-nout (4), roz-hrab-at (4), roz-hran-í (4), roz-hryz-at (2), pod-hrad-í (2)

1+

2

ode-pra-t (2)

10-

57

roz-proud-it (1), roz-prask-aný (9), od-prásk-nout (1), vz-pruh-a (3), bez-prašný (9), bez-práv-í (9), pÍed-prs-eÁ (4), od-prac-ovat (4), od-pro-dat (15), pod-prçm!rný (2)

1+

1

beze-sn-ý

1-

1

pod-sn!ñ-ník (1)

1+

2

vze-šl-ý (2)

2-

3

roz-šlap-at (2), roz-šleh-at (1)

1+

4

pode-zd-ít (4)

2-

3

od-zdol-a (1), nad-zdvihn-out (2)

+

4

beze-dn-ý (4)

dr

hr pr

sn šl zd dn

12

6 11

2 3 3 1

(20) counter-examples (one example per root followed by the number of items representing it in brackets; total number of contravening items: 73) s+C Hiatus syllabic C2 beze-sráñkový (1) od-frknout (4) ode-stÍít (5) od-chrchlat si (1) ode-stlat (6) beze-skvrnný (1) beze-slovný (1) roz-trñení (14) roze-štvat (1) roze-smát (1) roz-vrstvit (2) beze-sporný (1) roz-vrtat (3) roze-spalý (1) pod-vrh (14) beze-stopý (1) pod-hrnout (4) roze-znat (4) roz-vlnit (2) beze-zvu…ný (3) od-vlhnout (2) ode-dávna (1) (21) all counter-examples are expected to behave in this way a. in Hiatus position ...Cx - Cx..., vowels appear more generally in the language, cf. (22). b. syllabic consonants behave more generally like vowels, cf. (23), thus [CCC]=/CVC/. c. s+C clusters behave as one consonant elsewhere in the language, cf. (24). (22) Hiatus a. ...Cx+CxV... se šatnou ze šatny se silou ze síly ve válce ze zeleniny ke káv!

b. ...Cx+CyV... v šatn! k šatn! v síle k síle k válce k zelenin! v káv!

"avec/dans le vestiaire" "de/vers le vestiaire" "avec de la/dans la force" "de force/vers la force" "dans/vers la guerre" "de/vers des légumes" "au/dans le café"

(23) syllabic consonants behave like vowels Czech infinitives must be bimoraic: V+V d!lat VV krást V+Csyll trp!t inf 1E, 2E sg ind past active participle krás-t krad-u, krad-eš kradl rçs-t rost-u, rost-eš rostl krý-t kry-j-u, kry-j-eš kryl stá-t se stan-e se stal se zná-t zn-ám znal dlí-t dl-ím dlel prá-t per-u pral (24) s+C a. the incriminated roots bear a ø not as predicted /søtC/, but after C2 /støC/: stC_=pf st_C=ipf ode-stÍí-t po-stír-at ode-stl-at roz-stýl-at roze-štv-at po-štív-at

b. s+C clusters behave as a single consonant 1. radical V-zero alternations never over a CC *[CøCC-V], only exception: CC=st √Cest-ø √Cøst-V NOMsg GENsg lest løst-i "cunning" kÍest kÍøt-u "baptism" …est cøt-i "honour" 2. group-palatalisations only with s+C: adj. NOM sg adj. NOM pl mokÍÍ-í kr mokr-ý stv …erstv-ý …erstv-í h drah-ý draz-í ch hluch-ý hluš-í dobÍÍ-í br dobr-ý vs. sk …esk-ý …ešt-í ck historick-ý histori……t-í dl

infinitive past passive participle obydl-it obydl-en

vs.

sl st zd

infinitive mysl-et …ist-it jezd-it

past passive participle myšl-en …išt-!n jeññd-!n

Datation (25) [Prefix+Root] are lexicalized items stored as one in the lexicon. Prefixation is synchronically inactive. a. heavily restricted productivity. b. analogical activity: pf roze-mlít provokes roze-mílat analogically. c. Hiatus-situations are treated at random: e may appear or not. d. not any prefix can be combined with any stem. By contrast, any preposition can stand before any stem. e. in some cases (11 roots), the vowel appearing within #[CC] can be detected in former stages of the language only: roze-Ívat "begin to shout" has no related CVC-form synchronically. But Old Czech Modern Czech 1° Íev-u Ív-u 2° Íev-eš Ív-eš 3° Íev-e,... Ív-e (26) When did prefixation take place? When did Prefix and Root stop being two distinct lexical items? (27) incorporation of an affixal consonant into the stem IE *sul (lat sol) > CS *sXl-n-\ce > cz slunce od-sluní the above analysis predicts /CCvC/ in CS, the structure was /CvCC/ hence, /CvCC/ > /CCvC/ must have been achieved when prefixation took place. ==> the analysis predicts that prefixation took place after CS times.

(28) Slavic Metathesis IE *ghordh- >

C__LC lat h o r t us ger Ga r t en CS *g o rd-

CL__C >

OCS cz

gradX h r a d (rus gorod)

Roots having undergone META never bear the prefixal -e-: roz -blácený < psl *bol-to od -blanit < psl *bol-na pÍed-bÍeznový < psl *berza bez -hlavý < psl *gol-va roz -hlaholit se < psl *gol-gol roz -hlas < psl *gol-sX bez -královí < psl *karl bez -mraký < psl *mork roz -mrazit < psl *morz vz -planout < psl *pol-nút roz -plamenit < psl *pol-men roz -plašit < psl *polch

od -plazit se < psl *polzX od -plivnout < psl *pelnX vz -tlak < psl *tolk roz -trhat < psl *tXrg-ati roz -trnout < psl *t\rp-n pod -vrátit < psl *vert od -zbrojit < psl *borj roz -broj < psl *borj bez -dÍevý < psl *dervo roz -mlátit < psl *moltX bez -prašný < psl *porch od -pracovat < psl *port-ja

od -pÍedu < psl *per-d\ od -stÍedit < psl *serda < IE *kerd od -stranit < psl *stor-na bez -tÍídní < psl *…erda bez -vládí < psl *vold roz -vláknit < psl *volk bez -vlasý < psl *volsX od -vléci < psl *velk-ti roz -tlouci < psl *tolk-ti

absence of prefixal -e- implies a CL__C structure ==> the analysis predicts that prefixation took place after META. (29) yer-vocalization traditional view: three different origins of modern Slavic vowel-zero alternations. a. yers CS *d\n\ > cz den vz GEN døn-e CS *po-dX > cz pode-brat vs. podø-bradek b. Ablaut CS inf-stem *b\r-, finite-stem *ber > cz inf bør-át vs. finite forms ber-u,... c. epenthesis 1. feminine i-stems: NOMsg píseÁ-ø - GENsg písn-! < NOMsg psl *p!-sn\ NOMsg báseÁ-ø - GENsg básn-! < NOMsg psl *ba-sn\ < IE *bh~ 2. neuter o-stems: GENpl …ísel-ø - NOMsg …ísl-o < NOMsg psl *…it-sl-o < IE keit vs. GENpl psl *…it-sl-X GENpl sester-ø - NOMsg sestr-a < GENpl psl *sestrX 3. masculine o-stems: NOMsg mozek-ø - GENsg mozk-u < stsl NOMsg mozgX 4. prefixes/ prepositions such as cz vz(e), roz(e), bez(e), z(e), ot(e)/od(e) e.g. cz vz - vze (e.g. vze-pnout se - vz-pínat se)< stsl vXz- without final yer. a single statement covers all cases: any empty Nucleus escaping Proper Government (="in strong position") was subject to epenthesis. This concerns Nuclei formerly filled with yers as well as Nuclei that have always been empty. ==> prefixation must have occurred before the end of this epenthesis (="vocalisation of yers").

(30) consistent datation: no contradiction. summary datation of prefixation1 indicator

relative datation according to phenomena

relative: period

absolute: year

"slunce"

after s_l > sl_n

CS or later

500 BC or later

after META

after 825 AD

META

after Slavic Metathesis. Datation of META: late CS, before loss of yers before the end of yer-vocalisation

late CS, after META2

West: app. 1025 AD

yers

prefixation occurred in the 10th century

1

see classical philological evidence such as e.g.Lamprecht (1987), Panzer (1991), Arumaa (1964), Trávní…ek (1935), Vondrák (1906), Komárek (1962), Lamprecht et al. (1986), Gebauer (1894). words of the shape [CVLjerC] (L=liquid) such as kol\ce, volX…\kX become kolce, vol…ek, not **kloce, vlo…ek. Hence, META was not active anymore when yers fell out. 2

Relevance of Czech prefixal alternations (31) vowel-zero alternations cross-linguistically ("e" being an alternation-site) (Scheer 1997): zero CeC-V

vowel CeC-ø

vowel CeC-CV

gloss

Moroccan Arabic

kvtøb-u

køtvb-ø

kvttvb-ø

they have written, he has written, he has caused to write

German (optional elision)

xnø“-c

xnc“

xnc“-lxç

inner+infl, inner, internal

Tangale (Chadic)

dobø-go

dobe

dobu-n-go

called, call, called me

Somali (Cushitic)

nirøg-o

nirig-ø

nirig-ta

young female camel pl, sg. indef., sg. def.

Turkish

devør-i

devir-ø

devir-den

transfer ACC, NOM, ABL

Slavic (e.g. Czech)

lokøt-e

loket-ø

loket-ní

elbow GENsg, NOMsg, Adj.

podø-kova

─

podø-bradek

horseshoe, double chin

BUT Czech prefixes

(32) a. CeC1-C2 normally provokes the vowel. b. only in Czech prefixes, C1 and C2 are mono-morphemic. c. ==> the key to the extraordinary behaviour of Czech prefixes must be found in the relation C1 and C2 contract (cf. Scheer 1996). (33) current approaches to vowel-zero alternations: a. epenthesis as a consequence of syllabification, e.g. Wiese (1988). b. Government: "Proper Government cannot apply over governing domains", e.g. Kaye et al. (1990). (34) approaches relying on syllabification assume that lexical structures are non-syllabified. They are thus unable to encode the crucial lexical difference √ BRaD vs. √ BøR ==> lexical structures are fully specified for syllabic structure. (35) "Proper Government (PG) cannot apply over governing domains" is falsified: √BRaD "beard"

│ │ │

'(

│ │ │

> │ > │ │ └───┘ │ │ │ └─────────┘ PG

(23) alternative: CVCV syllable structure (Lowenstamm 1996) closed syllable O N O N ││││ C V C ø (27)

geminate O N O N \ / │ C V

long vowel O N O N \ / C V

│

PG ┌─────┐ ? │ │ │ │

│

│ │

│

(28) ┌────────────────┐ │ ┌ │┐ ? │ ││ │ │ │ │ │ ││ │ ││ │ │ │ ══ │ │ │ └ ┘ autonomous domain

APPENDIX Analogical activity (36) misbehaving items in regard of (2b) "only √CC-initial stems provoke prefixal -e": analogical activity. ...e-√CV: 21 items ...e-√CV related √CC- form of the same stem roze-mílat

roze-mlít

roze-sílat

roze-slat

roze-bírat

roze-brat

pode-mílat

pode-mlít

pode-zírat

pode-zÍívat (ipf)

ode-pírat

ode-pÍít

pÍede-sílat

ipf

pÍede-slat

pf

vze-jít

inf

vze-jdu

infl. form s

pÍede-jít

pÍede-jdu

roze-jít

roze-jdeme se

nade-jít

nade-jdu

pode-jít

pode-jdu

ode-jít

ode-jdu

roze-…íst

roze-…tu

pode-šev

N/As g

pode-šve

G sg

√síl

ode-slaní

roze-bíratelný

√bír

roze-brat

pode-zíravý

√zír

pode-zÍívat

ode-sílání ode-sílací ode-sílatel roze-bírací

(37) regular alternation pf=ø ipf=í u-søn-out u-sín-at vy-bír-at vy-bør-at po-søl-at po-síl-at po-štøv-at po-štív-at

"s'endormir" "choisir" "envoyer" "exciter, provoquer"

(38) regular

irregular

pf CC==> +e

ipf CVC==> -e

pf CC==> +e

ipf CVC==> -e

expected ipf: -e

vze-dmout

vz-dout

roze-mlít

roze-mílat

roz-

roze-tnout

roz-tít

roze-slat

roze-sílat

roz-

roze-psat

roz-pisovat

roze-brat

roze-bírat

roz-

nade-psat

nad-pisovat

pode-mlít

pode-mílat

pod-

ode-psat

od-pisovat

pode-zÍívat (ipf)

pode-zírat

pod-

roze-stlat

roz-stýlat

ode-pÍít

ode-pírat

od-

pode-pÍít

pod-pírat

pÍede-slat

pÍede-sílat

pÍed-

ode-pÍít

od-pírat

pode-jmout

pod-jímat

ode-brat

od-bírat

ode-hnat

od-hán!t

ode-mknout

od-mykat

ode-pnout

od-pínat

ote-vÍít

ot-vírat

Total regular items: 14

Total irregular items -e-√CVC: 8

References Arumaa, Peeter 1964. Urslavische Grammatik. Band I: Einleitung, Lautlehre (Teil I: Vokalismus, Teil II: Betonung). Heildelberg 1964. Gebauer, Jan 1984. Historická mluvnice jazyka …eského. Vol. I: Hláskosloví. Praha 21963. Kaye, J.D., J. Lowenstamm, J.-R. Vergnaud 1990. Constituent structure and government in phonology. In: Phonology Yearbook 7.2, 193-231. Komárek, Miroslav 1962. Historická mluvnice …eská. Volume I: Hláskosloví. Praha 31969. Lamprecht, Arnošt 1987. Praslovanština. Brno. Lamprecht, Arnošt, Dušan Šlosar, Jaroslav Bauer 1986. Historická mluvnice …eštiny. Praha. Lowenstamm, Jean 1996. CV as the only syllable type. In: Durand, Jacques, Bernard Laks (eds): Current trends in Phonology: Models and Methods, 419-441. Salford, Manchester. Panzer, Baldur 1991. Die Slavischen Sprachen in Gegenwart und Geschichte. Sprachstrukturen und Verwandtschaft. Frankfurt/Main. Scheer, Tobias 1996. A theory of direct consonantal interaction. Doctoral dissertation, Université Paris 7. Scheer, Tobias 1997. Vowel-zero alternations and their support for a theory of consonantal interaction. In: Bertinetto, P.M., L. Gaeta, G. Jetchev, D. Michaels (eds): Certamen Phonologicum III, 6788. Torino: Rosenberg & Sellier. Trávní…ek, František 1935. Historická mluvnice …eskoslovenská. Úvod, Hláskosloví a Tvarosloví. Praha. Ulbrich, Rolf 1978. Langenscheidts Taschenwörterbuch der tschechischen und deutschen Sprache. Erster Teil Tschechisch-Deutsch. Berlin, München, Wien, Zürich, New York 101993. Vondrák, Wenzel 1906. Vergleichende Slavische Grammatik. I.Band Lautlehre und Stammbildungslehre. Göttingen. Wiese, Richard 1988. Silbische und lexikalische Phonologie. Tübingen: Niemeyer.

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28.

29. 30. 31. 32. 33. 34. 35. 36. 37. 38.